MiSTer FPGA Forum

thisisamigaspeaking wrote: ↑Tue Dec 13, 2022 2:11 pm

Again I don't understand what the framework is in this context,

The framework is the sys/ subdirectory from this repository: https://github.com/MiSTer-devel/Template_MiSTer

It's the code that provides cores with a uniform menu with options, with the possibility to communicate with the main executable for receiving ROMs, disks, inputs and whatnot, and more.

In short, it's what makes an FPGA binary be a MiSTer core. There's a copy in every core including this one (except the Jotego ones which use a different framework, compatible with MiSTer's, and also GPL'ed).

Ok yeah, from my understanding of the GPL, you can't take those files and put them in a combined package under some other license, or use portions with incompatible licenses with them.

These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

I think the framework should ideally be BSD, MIT or Apache licensed. Not sure how LGPL would apply. GPL can be tricky.

But what is this paranoid tantrum?
Come on, please, keep stirring the hornet's nest.

I think I've read too much nonsense for today, if you don't like the core , just delete it and let others do cores, official or not.
Unfortunately, all you have to do is search the net to find out which foot you're limping on.: http://www.formauri.es/personal/pgimeno/spec/spec.html

The big difference is that the PDP2021 core in MiSTer only has a slight drawback, its author does not want the source to be published in Git since it is a Micro$oft company and as a good lover of free software, no water for those .

The core can be published on any site, which of course is not being owned by M$.

Not only do you get confused and tangled by yourself, but you are confusing others, first you clarify yourself and when you are ready then you remove the drama from your comments and everything will be normal.

And talking about life, what a coincidence:

https://github.com/DavidJRichards/pdp2011/
https://github.com/birdybro/PDP11_MiSTer

Thank you for your criticism and for your discouragement. You're doing a great job

Free licenses are just that, FREE

I want to think that omitting me from the template contributors was a mistake..... I don't have to ask. As I don't need the framework for this core. My mister has PS2 , VGA and sdcard. This was only a FREE try to have a mister PDP11 core. Bot don't worry. I'm going to delete this and change the port for de10nano.... I hope terasic is not going onto sue me for using their trademark. And sorry. Never had vulnerated a license. I have not vulnerated any license here. And I WILL NEVER VIOLATE A LICENSE WITH A MISTER CORE. full stop. And sorry to everybody.

ron wrote: ↑Tue Dec 13, 2022 3:06 pm

Free licenses are just that, FREE

MIT, BSD and Apache are permissive. GPL is not, and it is intended to protect from both commercial exploitation (some types) and "free" exploitation that places restrictions on previously "free" code. It's been called "viral" as well and that is part of the intent and what one is assuming and agreeing to when publishing GPL code.

I don't personally mind using something with pirated code in it, depending on context (if it's harmless and noncommercial), but that doesn't change what it is. What I would not do myself is distribute such code.

I don't think pgimeno is wrong for pointing out the provisions of GPL. There were assertions made about copyright (this code is copyrighted, you can only use it for certain things) and pgimeno contradicted that.

Saying "this code is so free you can't even use it for what you want (commercially)" doesn't sound very free. I can understand not wanting such work to be used commercially but I think it is a misunderstanding of licenses and it's a good idea to point out why. GPL and LGPL are the best way to release free software in most cases, unless a permissive license is necessary (as in the case we see here, an interface or glue type situation where it should be permissive). A CC-BY-NC license or whatever, I don't understand the point of that.

The GPL applies to software. The actual Verilog and VHDL code that comprises the cores describes hardware and is not covered under it.

dmckean wrote: ↑Tue Dec 13, 2022 4:34 pm

The GPL applies to software. The actual Verilog and VHDL code that comprises the cores describes hardware and is not covered under it.

Not sure I understand, you just described it as "code". Hardware is physical, Verilog and VHDL are not physical. It could certainly be licensed under whatever someone likes (perhaps a hardware license), but once combined with GPL-licensed code, why wouldn't it have to be GPL?

Is there some discussion of this or reference to look at online? I'm very curious now.

Just looking into it quickly it looks like none of the Verilog or VHDL anyone is using anywhere can be copyrighted in the US if they are considered schematics, they are not copyrightable. So anyone is free to use them (barring patents)? Is that correct?

thisisamigaspeaking wrote: ↑Tue Dec 13, 2022 4:50 pm

Just looking into it quickly it looks like none of the Verilog or VHDL anyone is using anywhere can be copyrighted in the US if they are considered schematics, they are not copyrightable. So anyone is free to use them (barring patents)? Is that correct?

Possibly.

May be another thread would be more appropriate. PDP11 MISTer is no more.
Thanks @rampa
Thanks for bringing up the GPL

rampa is fine distributing his core on his or her own without a GPL license. This is just like Ubuntu distributing non-GPLed NVidia drivers as a separate download after purchase.

However, if he or she ever wants it to be a default part of Mister, it may need to be GPLed.

Malor wrote: ↑Sun Dec 11, 2022 9:00 pm

oot user, make sure to add that user to the dialout group. If you're running as root, it should work automatically.

I think maybe ppp starts automatically, so you might need to disable that if you want to use SLiRP instead. I'm trying to figure out how ppp is started, so you can disable it.

second edit: after some experimentation, it looks like the Amiga core automatically launches the ppp daemon, and then closes it when it exits. Assuming there's no SLIP for the PDP2011 core to start, and that it's using the same allocation system that the Amiga is, then at least theoretically, you should be able to connect SLiRP to /dev/ttyS1. You should be able to launch it with /etc/inittab, but keep in mind that it will probably hold a lock on the port, so that you can't use PPP networking with other cores while it's running. And you may need to give it some kind of 'detach' argument to get it to run as a daemon, instead of from a user console as it was intended.

Anything that use the serial(PPP, console, MIDI, modem, etc..) is started by /sbin/uartmode which gets called every time a core starts. If the OSD UART option is net to none, then nothing will on /dev/ttyS1.

I would suggest reading altering uartmode script do do the following:

[ -f /tmp/CORENAME ] && corename=$(cat /tmp/CORENAME) 
if [ "$corename" == "PDP2011" ]; then
	--> do start SLIP/SLiRP stuff here
else
	--> start PPP normally
	taskset 1 pppd $conn_speed file /tmp/ppp_options
fi

Malor wrote: ↑Tue Dec 13, 2022 5:23 pm

rampa is fine distributing his core on his or her own without a GPL license. This is just like Ubuntu distributing non-GPLed NVidia drivers as a separate download after purchase.

However, if he or she ever wants it to be a default part of Mister, it may need to be GPLed.

I am not sure about that. It looks more likely it will be a regular DE-10 Nano core using PMODs (or DIY boards) for VGA, PS2 and SD card. In this case you could still run it on MISTer but you would have to ditch your IO board if you have one, use the previously mentioned boards. Once started it would have no menu and you would have to reboot you MISTer.
I may be wrong but it is what I understand.

I got the impression that pdp-11 for MiSTer does not really need anyone. And most of this thread is a discussion of licenses, not the PDP-11 itself. It is sad.

xolod79 wrote: ↑Tue Dec 13, 2022 8:59 pm

I got the impression that pdp-11 for MiSTer does not really need anyone. And most of this thread is a discussion of licenses, not the PDP-11 itself. It is sad.

It's very unfortunate because I would like to see this core. It was OP who brought up the issue of copyright though, and it had to be discussed at that point.

rampa wrote: ↑Tue Dec 13, 2022 3:18 pm

I want to think that omitting me from the template contributors was a mistake..... I don't have to ask.

I didn't omit anyone; I listed the ones I saw in the first page of commits and added "and others", as that was enough to make my point. I didn't know you're a contributor; if you're a copyright holder, that's great, as that's one less person to ask permission to, if you want to distribute code that includes the framework with a different license.

This whole conundrum might have a solution (besides asking every copyright holder to make an exception for this case), but it is tough. The idea would be to ask for permission from as many copyright holders of the framework as possible - permission for changing the license from GPL to a more permissive one. And then to filter the commits in order to remove code from those who can't be contacted, or who don't agree. Then build on that, with a different license. It's possible that features are dropped along the way, and that there's a need to reimplement them with new code not copyrighted by those who haven't given permission.

Anyone willing to post a proposal to relicense the framework in the Template_MiSTer repo? I can't because I left MSGitHub to never look back.

Edit: Oh, and of course there's another way, which is to ask the PDP module creator to reconsider licensing the code under a GPL-compatible license (like the GPL itself). Or even dual licensing, like Oracle does with MySQL for example. This option may look appealing to the PDP creator if they are well informed of the implications.
 

@ron:

ron wrote: ↑Tue Dec 13, 2022 3:06 pm

Free licenses are just that, FREE

You seem to be confusing "free" as in the freedom defined by the GPL with "free beer", just like you're confusing Git (a source control tool like Hg, SVN or CVS) with MSgitHUB (a web service which I left almost immediately after the buyout). This is the philosophy that inspired the GPL: https://www.gnu.org/philosophy/free-sw. ... r-freedoms - note "freedom 0" says "The freedom to run the program as you wish, for any purpose", not "for any purpose as long as it's not commercial".

From the two links you posted, one does not use GPL'd code for the core; the other does and may suffer from the same issues as the core that was published here. [Edit: No, I'm wrong. The one by birdybro uses a GPL PDP-11 module, not the Syste van Slooten one, and therefore the restriction of not using it for commercial use does not apply. That one appears to be truly free.]

Another thread for GPL related stuff PLEASE!

I want to express my gratitude to the author of this kernel! This is a job well done. The only thing I would like to see is ethernet which is brought to hps via the TAP interface in linux, but this will require a special driver.

Sooo, in short... someone ports a great core, another one comes with legal gibberish and we are f**** up. Great, I despise this nonsense. But, hey man, we are adults so keep it legal baby.

Regarding the core, great work @rampa. Don't let this discourage you for future developments.

P.S. This thread smells like teen spirit

xolod79 wrote: ↑Wed Dec 14, 2022 4:05 pm

I want to express my gratitude to the author of this kernel! This is a job well done. The only thing I would like to see is ethernet which is brought to hps via the TAP interface in linux, but this will require a special driver.

Good idea. I'll explore this way also. Is this it yet done on any core?

The "driver" is done in PDP assembly. so it is possible.

Only a note. Deleted the download from the forum (no "illegal" content on the forum, isnt it?), but the files on mega (referenced on the first post) are still there. Download at your own risk

@rampa: the core is great, I had a very good time with it the day you published it! I hope you continue working on it, please ignore stupid license issues. Don't let certain individuals waste your work: we all lose if they get their objetive. Simply ignore them: pretent they are not there. They don't deserve any attention.
I love old Unixes and we really need this core on the MiSTer!

@pgimeno: What the hell is wrong with you, man? I have no words, please go away and don't come back to any thread where I can read you, what a disgusting person you seem to be, really.

ron wrote: ↑Tue Dec 13, 2022 3:06 pm

Free licenses are just that, FREE

pgimeno wrote:

You seem to be confusing "free" as in the freedom defined by the GPL with "free beer", just like you're confusing Git (a source control tool like Hg, SVN or CVS) with MSgitHUB (a web service which I left almost immediately after the buyout). This is the philosophy that inspired the GPL: https://www.gnu.org/philosophy/free-sw. ... r-freedoms - note "freedom 0" says "The freedom to run the program as you wish, for any purpose", not "for any purpose as long as it's not commercial".

From the two links you posted, one does not use GPL'd code for the core; the other does and may suffer from the same issues as the core that was published here. [Edit: No, I'm wrong. The one by birdybro uses a GPL PDP-11 module, not the Syste van Slooten one, and therefore the restriction of not using it for commercial use does not apply. That one appears to be truly free.]

Yes, Of course, please, make sure that you add to the core that you publish the corresponding authors and rights that are in force today and start complying with them yourself from now on.

I have not seen in any of your repositories any note about it, just in case I remind you. It's one thing to have a GPL or GNU license from which nobody gets any benefit or profit. On the other hand, you use a manufacturer's ROM and you don't even respect the licensing they grant us.

Amstrad are happy for emulator writers to include images of our copyrighted code as long as the (c)opyright messages are not altered and we appreciate it if the program/manual includes a note to the effect that "Amstrad have kindly given their permission for the redistribution of their copyrighted material but retain that copyright".

with which it is free, but Amstrad maintains its copyright and license...

https://web.archive.org/web/20120509012 ... d-roms.txt

Could you tell us where the reference to Amstrad is in your repositories or emulators?

For GPL related stuff please post here:
viewtopic.php?p=65851#p65851

This current thread is already 2 pages and almost nothing on the pDP11 itself.

Thank you for your cooperation.

rampa wrote: ↑Wed Dec 14, 2022 7:41 pm

xolod79 wrote: ↑Wed Dec 14, 2022 4:05 pm

I want to express my gratitude to the author of this kernel! This is a job well done. The only thing I would like to see is ethernet which is brought to hps via the TAP interface in linux, but this will require a special driver.

Good idea. I'll explore this way also. Is this it yet done on any core?

The "driver" is done in PDP assembly. so it is possible.

'driver' I meant linux hps kernel driver. It is needed to transfer ethernet frames from fpga to hps and vice versa. Here is an example of such a diver.
https://github.com/scrameta/MiSTer_Hybr ... minimigirq
This driver will provide an ioctl that will be used by a small application in user space hps that will send frames to the TUN virtual device.

NinjaWarrior wrote: ↑Wed Dec 14, 2022 7:09 pm

Sooo, in short... someone ports a great core, another one comes with legal gibberish and we are f**** up. Great, I despise this nonsense. But, hey man, we are adults so keep it legal baby.

Regarding the core, great work @rampa. Don't let this discourage you for future developments.

P.S. This thread smells like teen spirit

A lot of the reason you have all this great stuff to play with is the GPL. The vast majority of the software on the Mister exists because it can build on all the GPLed software that came before it, like the Linux kernel and all the userspace utilities.

It's not "legal gibberish", it's important. Trying to stick proprietary code into a GPLed project can, in some cases, be outright illegal, at least to the extent that copyright infringement is illegal. Most GPL holders are pretty reasonable about stuff, generally only demanding that you live up to the license terms, but the disruption downstream can be considerable. Making sure everyone complies with the license terms up front can save a lot of hassle.

Had earlier projects not had similar discussions, enforcing the terms of the license, you would not have a MiSTer. Ignore thirty years of history at your peril.

xolod79 wrote: ↑Wed Dec 14, 2022 10:31 pm

rampa wrote: ↑Wed Dec 14, 2022 7:41 pm

xolod79 wrote: ↑Wed Dec 14, 2022 4:05 pm

I want to express my gratitude to the author of this kernel! This is a job well done. The only thing I would like to see is ethernet which is brought to hps via the TAP interface in linux, but this will require a special driver.

Good idea. I'll explore this way also. Is this it yet done on any core?

The "driver" is done in PDP assembly. so it is possible.

'driver' I meant linux hps kernel driver. It is needed to transfer ethernet frames from fpga to hps and vice versa. Here is an example of such a diver.
https://github.com/scrameta/MiSTer_Hybr ... minimigirq
This driver will provide an ioctl that will be used by a small application in user space hps that will send frames to the TUN virtual device.

The scheme I suggested earlier is redundant. Since the packets will first fall into the kernel space linux, then go to the user space of the linux application, and then they will return back to the kernel space TUN interface. It is more correct to make an ethernet card driver in linux with the minimum necessary functionality.

ron wrote: ↑Wed Dec 14, 2022 8:22 pm

Could you tell us where the reference to Amstrad is in your repositories or emulators?

I'm complying with Amstrad's requisites for permission by not altering any copyrights in the ROM that I'm distributing. "[W]e appreciate it if..." does not state a requisite, so I don't have an obligation to reference Amstrad. Anyway, please keep looking in my repositories, just in case I made any mistake, and report it if that's the case - I'll be glad to correct it to stay in compliance. Thanks in advance. Oh and please report them by PM instead of polluting the thread further, thanks.

I have uploaded the PDP core here:

https://github.com/MiSTer-Enhanced/PDP2011_MiSTer

In the future this repository will be synced to update_all. I won't enter into any discussion. If there is some legal problem report it to github.

antoniovillena wrote: ↑Thu Dec 15, 2022 2:29 pm

I have uploaded the PDP core here:

https://github.com/MiSTer-Enhanced/PDP2011_MiSTer

In the future this repository will be synced to update_all. I won't enter into any discussion. If there is some legal problem report it to github.

Great, thank you, PDP-11 is one of my most desired cores.

I don't think there was really any issue here at all. Apparently you can't even copyright electronics schematics (HDL) anyway, only patents would apply. OP made a statement that there was a restrictive copyright on the HDL and a discussion ensued. I don't think that's a reason for bad feeling on anyone's part.

Cheers all and happy holidays.